You are on page 1of 8

Analytical Chemistry Research 5 (2015) 18

Contents lists available at ScienceDirect

Analytical Chemistry Research


journal homepage: www.elsevier.com/locate/ancr

Direct electron transfer biosensor for hydrogen peroxide carrying


nanocomplex composed of horseradish peroxidase and
Au-nanoparticle Characterization and application to bienzyme
systems
Yusuke Okawa , Naoto Yokoyama, Yoshinori Sakai, Fumiyuki Shiba
Graduate School of Advanced Integration Science, Chiba University, 1-33 Yayoicho, Inage-ku, Chiba 263-8522, Japan

a r t i c l e

i n f o

Article history:
Received 4 March 2015
Accepted 18 May 2015
Available online 23 May 2015
Keywords:
Electrochemical biosensor
Direct electron transfer
Au-nanoparticle
Horseradish peroxidase
Glucose oxidase
Urate oxidase

a b s t r a c t
A reagentless electrochemical biosensor for hydrogen peroxide was fabricated. The sensor carries a
monolayer of nanocomplex composed of horseradish peroxidase and Au-nanoparticle, and responds to
hydrogen peroxide through the highly efcient direct electron transfer at a mild electrode potential
without any soluble mediator. Formation of the nanocomplex was studied with visible spectroscopy
and size exclusion chromatography. The sensor performance was analyzed based on a hydrodynamic
electrochemical technique and enzyme kinetics. The sensor was applied to fabrication of sensors for
glucose and uric acid through further modication of the nanocomplex-carrying electrode with the
corresponding hydrogen peroxide-generating oxidases, glucose oxidase and urate oxidase, respectively.
2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction
Recent progress in nanomaterials enables us new strategies in a
wide eld of science and technology. As nanomaterials feature
unique surface and electronic characters, they have attracted a
strong interest to apply them for fabricating functional systems,
of course, within the eld of biosensors [1]. Especially, nanoparticles are extensively studied in view of this [2]. For example,
nanoparticles as molecular labels, electrode catalysts, and nanoscale connectors between redox enzymes and base electrodes have
been reported.
We are interested in the electronic and surface characters of
nanoparticles as nanoscale interface for biofunctional molecules.
Several workers reported use of Au-nanoparticle (AuNP) for the
fabrication of electrochemical biosensors without soluble electron
mediators, and demonstrated that AuNPs are expected to present
an effective electrochemical interface to some enzymatic reactions
[36].
In a previous communication, we have briey reported a novel
high performance and reagentless electrochemical biosensor for
hydrogen peroxide using combination of horseradish peroxidase
(HRP) and AuNP [6]. The sensor is an indium-tin oxide (ITO)
Corresponding author. Tel.: +81 43 290 3452.

electrode carrying a monolayer of nanocomplex composed of an


AuNPs and HRP molecules, and is prepared in a very simple manner. The nanocomplex is immobilized through simply drop-casting
the nanocomplex dispersion on the electrode. The sensor gives
direct electrocatalytic current to H2O2 with high electron transfer
efciency at a mild potential without any soluble mediator.
Determination of hydrogen peroxide has been an important
topic in very wide elds, such as environmental, pharmaceutical,
clinical, and industrial applications, as reported in many works.
Our system requires the mild electrode potential comparable to a
soluble mediator system [7]. This is a strongly attractive feature
for minimizing non-specic electrochemical responses in practical
applications. Although there are several reports about the direct
electron transfer systems using the combination of the AuNP and
HRP [3,810], the working electrode potentials were generally
more negative, that is, they required a relatively high overpotential
for the direct electron transfer to HRP. Recent papers using the
combination of HRP and AuNP tend to hybrid approaches, that
is, the nanoparticles are combined with organic/inorganic materials [1114]. These approaches mainly aim to enhance the amount
of enzyme and/or functional site, and improve the sensitivity and
stability. Thus, the working potential of these sensors are not
improved relative to the conventional HRP/Au-nanoparticle systems reported. On the other hand, our system features not only

E-mail address: y_okawa@faculty.chiba-u.jp (Y. Okawa).


http://dx.doi.org/10.1016/j.ancr.2015.05.001
2214-1812/ 2015 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

simplicity but the extensively improved (i.e. mild) working


potential for H2O2 detection.
The improved working potential and high sensitivity for H2O2
detection are attractive to fabricate biosensors for oxidase
substrates. Oxidases generate H2O2 through the oxidation of the
substrates by dissolved O2. The sensitive detection of thus generated H2O2 gives sensors for the oxidase substrates by the combination with the oxidase.
In this paper we describe detailed analysis and further application of the above nanocomplex-based sensor. We rst describe
characterization of the nanocomplex and the immobilized layer.
We then discuss the kinetics of the surface process of the sensor
response of the nanocomplex-carrying electrode based on a hydrodynamic electrochemical technique and enzyme kinetics. For evaluating the applicability of the proposed sensitive H2O2 detection
system, we fabricated biosensors for glucose and uric acid by further immobilizing glucose oxidase (GO) and urate oxidase (UO)
onto the nanocomplex-carrying electrode, respectively.
2. Experimental
2.1. Materials
Horseradish peroxidase (HRP, EC1.11.1.7, Type VI), glucose oxidase (GO, EC1.1.3.4, from Aspergillus niger, Grade II), urate oxidase
(UO, EC1.7.3.3, from yeast) was purchased from Sigma, Boehringer
Mannheim, and Oriental Yeast, respectively. A series of pullulans
from Shodex was used as the molecular weight standards for size
exclusion chromatography (SEC). All other chemicals were of
reagent grade and used without further purication. Water was
puried with deionization and distillation. Indium-tin oxide (ITO)
coated glass plates (sheet resistance 10 X) and uoride-doped tin
oxide (FTO) coated glass plates (sheet resistance 10 X) were from
Geomatec and Nippon Sheet Glass, respectively.
2.2. Preparation of nanoparticles, nanocomplex, and electrodes
Quasi-monodisperse AuNPs with an average diameter of 16 nm
were prepared with the citrate reduction of 1 mmol L1 HAuCl4 at
the boiling temperature [1517].
A prescribed amount of HRP was dissolved at room temperature
in the AuNP dispersion prepared above. A prescribed amount of the
mixed solution was cast on the base electrode, and naturally dried
below 10 C to give an HRPAuNP nanocomplex-carrying electrode.
For the preparation of bienzyme electrodes, the nanocomplexcarrying ITO electrode was rst treated with 1% glutaraldehyde
solution for 15 min, rinsed with water, and then treated with the
enzyme solution for 15 min.
2.3. Methods
Electrochemical measurements were operated with the
conventional 3-electrode system with a Toho Giken potentiostat
Model 2000 and a Yokogawa Electric Works pen recorder.
Hydrodynamic experiments were performed with a rotating Au
disc electrode and a Hokuto Denko rotator. A home-made Ag/AgCl
electrode (KCl satd.) and a Pt wire were used as reference and counter electrodes, respectively. All experiments were done in a working
solution (a 0.1 mol L1 phosphate buffer of pH 6.4) at 25 C.
Transmission electron microscopic observation was done with
JEOL JEM-1200EX and Hitachi Denshi H-7650 electron microscopes.
UVVis absorption spectra were obtained with a Shimadzu
spectrometer Model UV3100.
Size exclusion chromatography was carried out with Shodex
Asahipak GS-620 columns (two columns in series, with a GS-2G

7B pre-column) and a Hitachi HPLC pump Model L-6000. The eluent was a 0.2 mol L1 phosphate buffer (pH 6.4) and the operation
was carried out at 50 C. The detection in a UV region was carried
out with a Hitachi diode array detector Model L-3000. The detection in a visible region was carried out with an Ocean Optics diode
array spectrophotometer Model USB2000 with a ow-through
optical cell.
The dc electrical conductivity of the immobilized layer was
measured using an Advantest digital multimeter Model R6871
and a comb-type electrode.

3. Results and discussion


3.1. Characterization of AuNPs and nanocomplex
Fig. 1(A) shows a transmission electron micrograph of the prepared AuNPs featuring the average diameter of 16 nm (17% CV).
The visible absorption spectrum of the dispersion was shown in
Fig. 1(B), exhibiting typical plasmon absorption at around 520 nm.
When a protein was dissolved in the dispersion, the red shift of
kmax was observed. Fig. 1(C) shows change in the kmax as a function
of the added HRP concentration. The mixed dispersion includes
6  1015 particles L1 of the AuNPs. As the enzyme concentration
increases, the kmax increases and shows saturation. The shift indicates the enzyme adsorption on the nanoparticle surface [18],
and the saturation of the shift suggests the saturation of the
enzyme adsorption.
Fig. 2 shows a typical size exclusion chromatogram for the
mixed dispersion of the AuNP and HRP. The dispersion contained
a sufcient concentration of HRP (20 lmol L1) to ensure the saturation of the enzyme adsorption. The detection was done with the
absorption at 210 nm. The trace shows three major peaks, A, B and
C. The peak C was assigned to citrate originally contained in the
reaction mixture from the retention time and the UV absorption
spectrum. The insets in Fig. 2 show the absorption spectra corresponding to the peaks A and B, respectively. The molecular weight
(pullulan equivalent) of the peak B calculated from the retention
time is reasonably agreed with the molecular weight of HRP
(44 kDa [19]). Correspondingly, the peak B exhibited absorption
maxima at around 210 nm (peptide bond), 280 nm (weak, aromatic amino acid residues), and 400 nm (heme) (see the inset).
The peak B corresponds to the (excess) free HRP molecules. The
peak A gives the absorption maximum at around 525 nm corresponding to the plasmon absorption of the AuNP, and light scattering towards the UV region. The pullulan equivalent molecular
weight of the peak A was 2.0  105, and hence, the corresponding
hydrodynamic diameter was estimated at 27 2 nm [20]. This
value is reasonable as the size of the AuNP (16 nm) with the
adsorbed HRP (5 nm [21]) monomolecular layer. We thus concluded that the peak A corresponds to the AuNPs with the
adsorbed enzyme monolayer. On the contrary, the AuNPs protected only with citrate (without the adsorbed protein) coagulate
and precipitate when the elution buffer is added to the dispersion.
The drastic improvement in the dispersion stability is also consistent with the enzyme adsorption.
In general, adsorption behaviors should be essentially discussed
through the adsorption isotherm. In Fig. 2, however, we could not
evaluate the real equilibrium (non-adsorbed or free) enzyme
concentrations in the ordinate. We attempted the determination
of the free enzyme concentration in the dispersion containing
6.0  1015 particles L1 of the AuNPs and 0.15 lmol L1
(9.2  1016 molecules L1) HRP as follows. The AuNPs were separated from the dispersion and the supernatant was concentrated
and analyzed with HPLC (the same setup with the above size exclusion chromatography). The detection limit of the enzyme through

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

(A)

(B)

0.35
0.30

Absorbance

0.25
0.20
0.15
0.10
0.05
0.00

400

450

500

550

600

650

700

Wavelength / nm

(C)
527

Wavelngth / nm

526
525
524
523
522
521

100 nm

520
0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

HRP concentration / mol L-1

Fig. 1. TEM image (A) and visible absorption spectrum (B) of the AuNP (as-prepared). (C) Change in peak wavelength of plasmon absorption of AuNP with HRP concentration.

this entire process was 10 pmol, and we could not detect any
enzyme from the supernatant. This means that more than 95% of
the added enzyme molecules were adsorbed at this enzyme concentration. We thus calculated that 15 molecules of HRP were
adsorbed on each AuNP at this condition. Below this added amount
level, we can consider the approximately complete adsorption of
the enzyme added on the AuNPs. When the enzyme concentration
increased, we can detect reasonable amounts of HRP with the
above method. The saturated adsorbed amount is estimated at
50 molecules/particle, being geometrically consistent with the
respective sizes of the AuNP and HRP molecule.
The monolayer adsorption of an enzyme on a nanoparticle
means the formation of a new functional nanocomplex. In the following sections, we discuss the characteristics of the electrodes
carrying this nanocomplex.

3.2. Sensor performance and preparative conditions


The above HRP/Au-nanocomplex was easily immobilized on an
ITO electrode surface by the simple drop-cast technique. The
deposited nanocomplex was sufciently stable against rinsing with
water. The nanocomplex-carrying electrode shows an excellent

100
ITO

Current / Acm-2

Fig. 2. Size-exclusion chromatographic trace of a HRP/AuNP mixed dispersion. The


insets indicate a UVVis spectra at the corresponding peaks.

sensor performance to H2O2 without any soluble mediator. We further examined the effect of the base electrode materials on the sensor characteristics using ITO, FTO, and Au as the base electrode.
Fig. 3 shows calibration graphs for the nanocomplex-carrying
electrodes of various base electrode materials. All electrodes
responded to the addition of H2O2 within several seconds under
continuous gentle stirring. The ratio of enzyme/AuNP (number
base) was 15, and the immobilized density of the nanocomplex
was 2.7  1011 particles cm2 (the nominal value for the immobilization process). The conditions were the optimized ones based
on the investigation described below. The electrode was operated
at +0.15 V vs. Ag/AgCl, where the direct electrochemistry of H2O2
is not observed for bare and AuNP deposited electrodes. The solution is continuously stirred at 1 rps and the response time is
rather limited with the stirring itself.
The sensor performance depended on the base electrode materials especially in sensitivity, although the current-concentration
proles were geometrically similar in essence. The sensitivity in
the case of an FTO base electrode was obviously the lowest among
these base electrode materials. It was also found that the
nanocomplex was apt to partially peel off from the FTO surface.
On the Au electrode, the nanocomplex was stably immobilized

Au

10

FTO
1

0.1

10

100

[H2O2] / mol L -1
Fig. 3. Typical sensor response of the nanocomplex-carrying electrodes of various
materials at +0.15 V vs. Ag/AgCl under continuous stirring (1 cps); j, ITO; d, Au;
N, FTO. HRP/AuNP ratio = 15:1, immobilized density = 2.7  1011 cm2.

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

but the sensitivity as the sensor was about half of the ITO electrode. These observations strongly suggest that the nanocomplex
effectively connected to the base electrode contributes to the sensor response, and that the contact between the AuNP core of the
nanocomplex and the base electrode surface is essentially important for the sensitivity. These results will be discussed in view of
the kinetic analysis below. ITO is the best choice for the present
nanocomplex system among the electrode materials examined.
Fig. 4(A) shows the sensor response as a function of the
HRP/AuNP ratio for ITO electrodes carrying the nanocomplex. The
curve shows a maximum at the ratio 15. When the ratio is larger
than 20, we observed that some of the deposited nanocomplexes
peeled off from the electrode. In this region, non-adsorbed enzyme
molecules are simultaneously deposited on the electrode. These
molecules do not strongly interact with the surfaces of both the
AuNPs and the base electrode, but lie between the nanocomplexes
and/or between the nanocomplex and the base electrode. When
these excess molecules dissolve, a part of the nanocomplexes
simultaneously peel off, and hence the sensor response decreases.
When the ratio is smaller, the limited amount of the introduced
enzyme limits the response. From the stability of the deposited
layer and the loaded enzyme amount, we chose the ratio of 15 as
the optimum.
Fig. 4(B) shows the sensor response as a function of the loaded
density of the nanocomplex on the ITO electrode. In the low density region, the response increases with the increase of the loaded
density, but saturates in the high density region. When the loaded

amount exceeds a certain value, the nanocomplexes exist as the


multilayer. When there is no electrical connection between the
nanocomplexes, only the nanocomplex in the rst layer can electrically connected to the base electrode, and can contribute to the
electrochemical response. We measured the electrical conductivity
of the nanocomplex layers deposited on a comb-type electrode.
The monolayer and multilayer of the nanocomplexes is highly
insulating (at least 1010 X of the sheet resistance) in a dry nitrogen
atmosphere, although a wet air atmosphere gives a measurable
conductivity originated by water adsorbed on the protein molecules. We thus concluded that there is no electrical connection
between the immobilized nanocomplexes.
Fig. 5 shows a TEM image of the deposited nanocomplex layer of
the optimized conditions. The lm was peeled off from the electrode
for the TEM observation. The particle density of 2.2  1011 cm2 was
observed from the photographs and reasonably agreed with the
loaded density of 2.7  1011 cm2. In this image, we can see that
the AuNPs do not contact with each other. This is caused by the
adsorbed protein layer (just visible as slight shadow in the TEM
images), and is consistent with the results of the electrical conductivity measurement and the HPLC analysis above.
In the immobilized nanocomplex layer, the AuNP cores do not
contact with each other, and electrical connection between the
nanocomplexes is negligible with this system. The multilayer loading of the nanocomplex is thus ineffective for the higher output
current because the nanocomplexes in the lowest layer can only
have the electrical connection to the base electrode. As seen in

30

(B)
Current / A cm-2

Current / A cm-2

25

10

(A)

20
15
10

0.1

5
0

10

100

HRP/Au-nanoparticle ratio

1000

0.01

10

10

11

10

12

10

-2

Density of nanocomplex / cm

Fig. 4. Sensor response as the functions of (A) the HRP/AuNP ratio (immobilized density = 2.7  1011 cm2; H2O2 concentration = 100 mmol L1) and (B) the immobilized
density (right, HRP/Au-nanocomplex ratio = 15:1; H2O2 concentration = 10 mmol L1). Base electrode, ITO.

500 nm

100 nm

Fig. 5. TEM images of coagulated nanocomplex lm peeled off from the electrode. Loaded density = 2.7  1011 cm2.

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

Fig. 4(B), the response is saturated at 5  1011 cm2 level. This


corresponds to the completion of the packed monolayer, being
consistent with the TEM observation in Fig. 5.
The stability of the electrodes was examined. The electrodes
were stored in air under refrigeration below 10 C when not in
use, and the response to 10 lmol L1 H2O2 was measured daily.
The sensor outputs after the 3 d storage were about 70% of the initial one. The electrodes show good reproducibility and stability
under these conditions. However, after the electrode was once
used in a high concentration of H2O2 (e.g. 1 mmol L1), the
response was suppressed to 5070% of the original level even when
it was again operated in a dilute H2O2 solution. When the electrode
is operated in such high H2O2 concentration conditions, the current
response also gradually decreased. This instability is due to the
intrinsic nature of HRP [22], and is an essential problem of
HRP-based biosensors. However, use of the electrode in these high
concentration conditions is a rather special case in bioanalytical
applications, and the solution of this problem is beyond the scope
of the present work.

the above analysis. This gure is an ideal calibration graph under


the conditions where the mass transfer is innitely rapid and only
the surface reaction is rate-determining. Based on this gure, we
can analyze the kinetics of the present system.
The HRP reaction with a soluble mediator is schematically represented as follows [23]:
k1

HRP H2 O2 ! Compound I H2 O
k2

Compound I Medred ! Compound II Medox


k3

Compound II Medred ! HRP Medox

1
2
3

here k1, k2, and k3 are rate constants, Med(red) and Med(ox) represent reduced and oxidized form of a soluble mediator, respectively,
and Compounds I and II are the intermediate states of HRP, respectively. The kinetic equation of the overall reaction was derived as
follows [7,24].

1
k2 k3

k1 H2 O2  k2 k3 Med

3.3. Hydrodynamic electrochemistry and kinetic analysis

The present electrode gave a steady-state current to the substrate when the solution was continuously stirred. When the stirring stopped, the current gradually decreased, and when the
stirring restarted, the current recovered to the original level. The
result means that the substrate supply to the electrode is one of
rate limiting steps in the sensor response.
We should analyze the surface process to characterize the
behavior of the nanocomplex separately from the mass transfer
process. We applied a hydrodynamic technique, that is, the rotating disc electrode (RDE) method to evaluate separately these processes. In this experiment, an Au disc electrode was employed
because of the availability of the electrode whose geometry is suitable for the RDE experiment. The rotation speed dependency of the
sensor response was examined in the electrolyte containing various concentrations of the substrate. The sensor output clearly
increased as the rotation speed increased. Fig. 6(A) shows
KouteckyLevich plots of the output current at various substrate
concentrations. Each plot gives a reasonable straight line with a
certain value of the intercept. This means the mass transfer of
the substrate and the surface charge transfer process are both
the rate determining steps. The slope of each plot gives an essentially identical diffusion constant of 7.1  106 cm2 s1 that is reasonable for the one of H2O2.
The intercept of KouteckyLevich plot gives the net charge
transfer rate without the mass transfer effect and we can exactly
evaluate the kinetics of the surface process. Fig. 6(B) shows a typical H2O2 dependence of the surface reaction rate determined by

here m is the rate of the overall enzymatic reaction. In the present


system containing no soluble mediator, the HRP intermediates,
Compounds I and II, must accept electrons from the electrode
through the AuNP core. Thus, the second term in the right hand
of Eq. (4) is a constant depending on the electrode potential.
Then, at a given electrode potential, Eq. (4) is redrawn as follows:

1
1

k1 H2 O2  mmax

here mmax is a parameter including the rate constants of charge


transfer between the HRP intermediates and the electrode, and represents the maximum rate under the condition of the sufcient substrate concentration. In addition, as the reactions take place only on
the electrode, the observed surface reaction rate V in Fig. 6(B)
depends on the surface density of HRP, CE, according to the following relationship:

V v CE

In Fig. 6(B), the experimental data reasonably obey on the best


t curve (solid curve) based on the above formulations. From this
analysis, we can calculate the maximum turnover as 30 s1 using
CE = 4.5  1012 molecules cm2 (estimated from preparative
conditions).
As described in the foregoing section, the current response
depended on the base electrode materials. Taking into account
the above analysis using an Au disc electrode, we estimated the
turnover of 70 s1 for the case of the ITO base electrode. This value

Fig. 6. Kinetic analysis using a hydrodynamic method. (A) KouteckyLevich plots for the nanocomplex carrying Au-disc electrode at various H2O2 concentrations. (B) Surface
reaction rate determined from (A) as a function of H2O2 concentration.

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

is order-of-magnitude higher than those of the related systems


reported in the literatures [3,25].
We can discuss the effect of the base electrode materials on the
sensor performance based on the above analysis. Returning to
Fig. 3, all the curves bend to saturate in a similar H2O2 concentration region, although the responses are linear in a lower H2O2 concentration range and the sensitivity of the electrodes differs.
According to the above kinetic analysis, this means that the specic
activity of the enzyme itself is similar and only the effective
amount of the enzyme differs among these electrodes. As the composition and the loaded density of the nanocomplex are identical in
Fig. 3, the apparent difference in the enzyme amount is caused by
the difference in the number of the nanocomplex that is electrically connected to the base electrode, that is, some parts of the
nanocomplex immobilized on an electrode do not work to be
called dead. Among the materials examined in this study, ITO
is the best choice for minimizing the dead nanocomplexes, however, further exploration of the electrode materials or chemical
modication of the electrode surfaces may improve this factor.
The turnover frequencies above were calculated based on the
loaded amount of HRP. However, taking into account the existence
of the dead nanocomplexes, the true specic turnover for the
working enzyme molecules may be higher, even in the case of
an ITO base electrode.
3.4. Working potential
We again stress the mild working potential of the present system. We examined the working potential of the present system
with cyclic voltammetry and the catalytic current measurement
at constant potentials. The catalytic currents were observed from
ca. +0.25 V vs. Ag/AgCl. Although the magnitude of the catalytic
currents increased at more negative potentials, the background
currents also increased at the more negative potentials. In view
of both the catalytic sensitivity and the background stability, we
concluded that the optimum potentials range from +0.15 to
+0.20 V vs. Ag/AgCl. This potential range is similar to that of the
electrode chemically immobilized HRP operated with ferrocene
derivatives as the soluble mediator [7]. On the other hand, Xiao
et al. [3] reported the catalytic direct electron transfer current
through immobilized AuNP/adsorbed HRP system at 0.30 V vs.
SCE, and Jia et al. [8] reported the working potential of 0.25 V
vs. Ag/AgCl of the electrode carrying a silica gel network incorporating AuNPs and HRP. Ruzgas et al. [25] reported the catalytic current at 0.00 V vs. SCE for an HRP-modied graphite electrode.
Comparing the potentials for these related systems, the present
system features notably positive working potential for the effective
electron transfer.
In the soluble mediator systems, Compounds I and II are
reduced by the soluble mediator (electron donor) according to

the above kinetic scheme (Eqs. (2) and (3)). Hexacyanoferrate(II)


and ferrocene derivatives are well-known as the electron donor
for HRP system [26]. The redox potential of these mediators is
about +0.2 V vs. Ag/AgCl in neutral pH conditions, and the biosensor using the combination of HRP and these soluble mediators
works at the potentials around this range [7,24,26]. In the present
system, the sensor works at the similar potentials, and the direct
electron transfer from the AuNP to Compounds I and II is as efcient as the typical soluble mediators. On the other hand, most of
HRPAuNP systems reported work at much negative potential such
as 0.2 V vs Ag/AgCl [3,810]. Similar working potentials for the
H2O2 detection were reported for an HRP-based nanowire system
[27], nonenzymatic electrocatalytic systems [2832], and
nanomaterial-based systems with non-HRP proteins [13,3337].
In these cases, it is reasonable to consider that the catalytic redox
centers do not work according to the above scheme. In most of
reported cases, H2O2 detection through direct electrochemistry of
HRP (and related electrocatalysts) is not based on the native reaction of HRP, but another scheme utilizing Fe(II)-state of heme as
demonstrated by the redox behavior of peroxidases [38,39].
3.5. Application to bienzyme systems
In spite of many reports about electrochemical biosensors using
the HRPAuNP combination, the present system features a noteworthy mild working potential, that is, the nanocomplex carrying
electrode works at very mild potentials with minimized interference from co-existing electroactive species in practical samples.
The present system is very suitable for monitoring the reaction
by H2O2-generating enzymes oxidases [7,26]. In view of this,
two oxidases, glucose oxidase (GO) and urate oxidase (UO), were
further immobilized on the nanocomplex-carrying electrode to
fabricate sensors for glucose and uric acid, respectively. These
enzymes catalyze the reaction between the corresponding substrate and dissolved oxygen to generate H2O2,

Substrate O2 ! Oxidized substrate H2 O2

Fig. 7 shows the calibration curves of the sensors for glucose


and uric acid. The working buffer was an air-saturated pH 6.4
phosphate buffer throughout, and all the operations were conducted at 25 C under continuous stirring (1 rps). The both electrodes reasonably responded to the corresponding substrates,
respectively, and gave the steady-state current within 30 s.
The linear sensitivity of the glucose sensor in the low concentration region is in 0.1 lA mM1 cm2 level and is order-of-magnitude
higher than that of 50 nA mM1 cm2 level for the electrode carrying HRP/GO bilayer operated with soluble mediators, that works
with the identical principle [7]. The immobilized amounts of GO
of the both sensors are in monolayer-modication level and estimated as in the same order. The sensitivity of the present system

0.3

3.0

Current / Acm-2

Current / Acm-2

2.5
2.0
1.5
1.0
0.5
0.0

20

40

60

80

100

120 140

Glucose concentration / mmol L

-1

0.2

0.1

0.0

50

100

150

200

250

300

Urate Concentration / mol L-1

Fig. 7. Response of the bienzyme systems. Left, glucose oxidase; right, urate oxidase.

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

is based on the highly sensitive detection of H2O2 generated by the


GO reaction. This is caused by the lack of loss of the Med(ox) reduction by diffusion to the solution bulk, and is a typical merit of
non-mediator systems [40,41].
Although the detection of H2O2 is popular for GO-based electrochemical biosensors [7,4244], the direct electrochemical detection
of H2O2 generated by the UO reaction is generally difcult. H2O2
generated by the UO reaction readily forms a complex with allantoin, another product of the UO reaction [45], and the detection of
free H2O2 is in principle limited [46]. Furthermore, the electrooxidation of H2O2 requires in general relatively high overpotentials
and we can hardly avoid the direct electrooxidation of uricate itself
that simultaneously occurs in the similar potential region. In view
of this, Nanjo and Guilbault [46] employed electrochemical monitoring of O2 consumption by the UO reaction for their historical
UO-immobilized platinum electrode. However, the primary products of the UO reaction are reported as 5-hydroxyisourate and
H2O2 [47], and a nonenzymatic post-reaction gives allantoin.
Rapid and sensitive detection of primarily generated H2O2 may give
a biosensor based on the detection of H2O2 just generated.
Electrochemical detection of UO-generated H2O2 has been studied
using several electrocatalysts [30]. Use of H2O2-specic enzyme
may also dissolve this problem. Tatsuma et al. [26] employed an
HRP/UO bilayer-modied electrode with a soluble mediator, and
successfully detected the mediated reduction current of H2O2 generated by UO, however, the linear sensitivity in the low concentration region is limited to ca. 50 nA mM1 cm2. The present
nanocomplex-based urate sensor exhibited the sensitivity of ca.
3 lA mM1 cm2 in spite of the comparable amount of the immobilized enzyme, conrming order-of-magnitude higher detection efciency of the present nanocomplex electrode.
3.6. Future scope
The present system demonstrates the facilitated electron transport between HRP and the base electrode through the AuNP as the
nanoscale connector. This is probably related strongly to the
adsorbed state of the enzyme molecules on the AuNP and some
characteristic electronic interaction between the AuNP and the
enzyme redox center (heme). In view of this, the examination of
the size dependence of the AuNP on the sensor performance may
give valuable information about the origin of the excellent charge
transfer character of the present system.
It is also noteworthy that our system employs
enzyme-nanoparticle nanocomplex that is spontaneously formed
(a kind of self-assembly, in a way) as a functional unit. The
nanocomplex can be easily immobilized to the electrode surface
of many kinds of substrate materials and potentially compatible
to some microfabrication technologies, such as ink-jet printing
the nanocomplex can be printed with an ink-jet printer at desired
point or area, and applicable to fabricating printed devices.
Acknowledgement
We thank D. Fujii, N. Kouzaka, M. Nakagomi, and A. Sagara
(Chiba Univ.) for their experimental assistance in part.
References
[1] J. Wang, Nanomaterial-based electrochemical biosensors, Analyst 130 (2005)
421.
[2] X. Luo, A. Morrin, A.J. Killard, M.R. Smyth, Application of nanoparticles in
electrochemical sensors and biosensors, Electroanalysis 18 (2006) 319326.
[3] Y. Xiao, H.-X. Ju, H.-Y. Chen, Direct electrochemistry of horseradish peroxidase
immobilized on a colloid/cysteamine-modied gold electrode, Anal. Biochem.
278 (2000) 2228.

[4] Y. Xiao, F. Patolsky, E. Katz, J.F. Hainfeld, I. Willner, Plugging into Enzymes:
nanowiring of redox enzymes by a gold nanoparticle, Science 299 (2003)
18771881.
[5] Y. Okawa, Y. Sakai, F. Shiba, Reagentless H2O2 biosensor with high electron
transfer efciency carrying nanocomplex of horseradish peroxidase and Aunanoparticle, Electrochemistry 76 (2008) 522524.
[6] K. Murata, M. Suzuki, K. Kajiya, N. Nakamura, H. Ohno, High performance
bioanode based on direct electron transfer of fructose dehydrogenase at gold
nanoparticle-modied electrodes, Electrochem. Commun. 11 (2009) 668671.
[7] T. Tatsuma, Y. Okawa, T. Watanabe, Enzyme monolayer- and bilayer-modied
tin oxide electrodes for the determination of hydrogen peroxide and glucose,
Anal. Chem. 61 (1989) 23522355.
[8] J. Jia, B. Wang, A. Wu, G. Cheng, Z. Li, S. Dong, A method to construct a thirdgeneration horseradish peroxidase biosensor: self-assembling gold
nanoparticles to three-dimensional solgel network, Anal. Chem. 74 (2002)
22172223.
[9] S.-Q. Liu, H.-X. Ju, Renewable reagentless hydrogen peroxide sensor based on
direct electron transfer of horseradish peroxidase immobilized on colloidal
gold-modied electrode, Anal. Biochem. 307 (2002) 110116.
[10] Q. Xu, C. Mao, N.-N. Liu, J.-J. Zhu, J. Sheng, Direct electrochemistry of
horseradish peroxidase based on biocompatible carboxymethyl chitosangold nanoparticle nanocomposite, Biosens. Bioelectron. 22 (2006) 768773.
[11] D. Shan, Q.-B. Li, S.-N. Ding, J.-Q. Xu, S. Cosnier, H.-G. Xue, Reagentless
biosensor for hydrogen peroxide based on self-assembled lms of horseradish
peroxidase/laponite/chitosan and the primary investigation on the inhibitory
effect by sulde, Biosens. Bioelectron. 26 (2010) 536541.
[12] M. Mathew, N. Sandhyarani, A novel electrochemical sensor surface for the
detection of hydrogen peroxide using cyclic bisureas/gold nanoparticle
composite, Biosens. Bioelectron. 28 (2011) 210215.
[13] N. Wei, X. Xin, J. Du, J. Li, A novel hydrogen peroxide biosensor based on the
immobilization of hemoglobin on three-dimensionally ordered macroporous
(3DOM) gold-nanoparticle-doped titanium dioxide (GTD) lm, Biosens.
Bioelectron. 26 (2011) 36023607.
[14] J. Shen, X. Yang, Y. Zhu, H. Kang, H. Cao, C. Li, Gold-coated silica-ber hybrid
materials for application in a novel hydrogen peroxide biosensor, Biosens.
Bioelectron. 34 (2012) 132136.
[15] J. Turkevich, P.C. Stevenson, J. Hillier, A study of the nucleation and growth
processes in the synthesis of colloidal gold, Discuss. Faraday Soc. 11 (1951)
5575.
[16] J. Turkevich, P.C. Stevenson, J. Hillier, The formation of colloidal gold, J. Phys.
Chem. 57 (1953) 670673.
[17] G. Frens, Controlled nucleation for the regulation of the particle size in
monodisperse gold suspensions, Nature 241 (1973) 2022.
[18] D. Eck, C.A. Helm, N.J. Wagner, K.A. Vaynberg, Plasmon resonance
measurements of the adsorption and adsorption kinetics of a biopolymer
onto gold nanocolloids, Langmuir 17 (2001) 957960.
[19] K.G. Welinder, Amino acid sequence studies of horseradish peroxidase. Amino
and carboxyl termini, cyanogen bromide and tryptic fragments, the complete
sequence, and some structural characteristics of horseradish peroxidase C, Eur.
J. Biochem. 96 (1979) 483502.
[20] C. Viebke, P.A. Williams, The inuence of temperature on the characterization
of water-soluble polymers using asymmetric ow eld-ow-fractionation
coupled to multiangle laser light scattering, Anal. Chem. 72 (2000) 38963901.
[21] T. Vegge, F.. Winther, B.R. Olsen, Horseradish peroxidase in plasma studied
by gel ltration, Histochemie 28 (1971) 1622.
[22] M.B. Arnao, M. Acosta, J.A. del Ro, R. Varn, F. Garca-Cnovas, A kinetic study
on the suicide inactivation of peroxidase by hydrogen peroxide, Biochim.
Biophys. Acta 1041 (1990) 4347.
[23] H. Yamada, I. Yamazaki, Proton balance in conversions between ve
oxidationreduction states of horseradish peroxidase, Arch. Biochem.
Biophys. 165 (1974) 728738.
[24] Y. Okawa, M. Nagano, S. Hirota, H. Kobayashi, T. Ohno, M. Watanabe, Tethered
mediator biosensor. Mediated electron transfer between redox enzyme and
electrode via ferrocene anchored to electrode surface with long
poly(oxyethylene) chain, Biosens. Bioelectron. 14 (1999) 229235.
[25] T. Ruzgas, L. Gorton, J. Emnus, G. Marko-Varga, Kinetic models of horseradish
peroxidase action on a graphite electrode, J. Electroanal. Chem. 391 (1995) 41
49.
[26] T. Tatsuma, T. Watanabe, Oxidase/peroxidase bilayer-modied electrodes as
sensors for lactate, pyruvate, cholesterol and uric acid, Anal. Chim. Acta 242
(1991) 8589.
[27] M.-J. Song, S.W. Hwang, D. Whang, Amperometric hydrogen peroxide
biosensor based on a modied gold electrode with silver nanowires, J. Appl.
Electrochem. 40 (2010) 20992105.
[28] X. Gao, L. Jin, Q. Wu, Z. Chen, X. Lin, A nonenzymatic hydrogen peroxide sensor
based on silver nanowires and chitosan lm, Electroanalysis 24 (2012) 1771
1777.
[29] W. Zhao, H. Wang, X. Qin, X. Wang, Z. Zhao, Z. Miao, et al., A novel
nonenzymatic hydrogen peroxide sensor based on multi-wall carbon
nanotube/silver nanoparticle nanohybrids modied gold electrode, Talanta
80 (2009) 10291033.
[30] W. Wang, C. Qin, Q. Xie, X. Qin, L. Chao, Y. Huang, et al., Rapid electrodeposition
of a gold-Prussian blue nanocomposite with ultrahigh electroactivity for dualpotential amperometric biosensing of uric acid, Analyst 139 (2014) 2904
2911.

Y. Okawa et al. / Analytical Chemistry Research 5 (2015) 18

[31] C.M. Welch, C.E. Banks, A.O. Simm, R.G. Compton, Silver nanoparticle
assemblies supported on glassy-carbon electrodes for the electro-analytical
detection of hydrogen peroxide, Anal. Bioanal. Chem. 382 (2005) 12
21.
[32] M.R. Guascito, E. Filippo, C. Malitesta, D. Manno, A. Serra, A. Turco, A new
amperometric nanostructured sensor for the analytical determination of
hydrogen peroxide, Biosens. Bioelectron. 24 (2008) 10631069.
[33] W. Yang, Y. Li, Y. Bai, C. Sun, Hydrogen peroxide biosensor based on
myoglobin/colloidal gold nanoparticles immobilized on glassy carbon
electrode by a Naon lm, Sens. Actuators B Chem. 115 (2006) 4248.
[34] C.-Y. Liu, J.-M. Hu, Hydrogen peroxide biosensor based on the direct
electrochemistry of myoglobin immobilized on silver nanoparticles doped
carbon nanotubes lm, Biosens. Bioelectron. 24 (2009) 21492154.
[35] A. Safavi, F. Farjami, Hydrogen peroxide biosensor based on a myoglobin/
hydrophilic room temperature ionic liquid lm, Anal. Biochem. 402 (2010) 20
25.
[36] C. Liu, J. Hu, Direct electrochemistry of hemoglobin entrapped in composite
electrodeposited chitosan-multiwall carbon nanotubes and nanogold particles
membrane and its electrocatalytic application, Electroanalysis 20 (2008)
10671072.
[37] S. Liu, Z. Dai, H. Chen, H. Ju, Immobilization of hemoglobin on zirconium
dioxide nanoparticles for preparation of a novel hydrogen peroxide biosensor,
Biosens. Bioelectron. 19 (2004) 963969.
[38] H.A. Harbury, Oxidationreduction potentials of horseradish peroxidase,
J. Biol. Chem. 225 (1957) 10091024.

[39] C.W. Conroy, P. Tyma, P.H. Daum, J.E. Erman, Oxidationreduction potential
measurements of cytochrome c peroxidase and pH dependent spectral
transitions in the ferrous enzyme, Biochim. Biophys. Acta Protein Struct.
537 (1978) 6269.
[40] T. Tatsuma, T. Watanabe, Y. Okawa, Model analysis of enzyme monolayer- and
bilayer-modied electrodes: the transient response, Anal. Chem. 64 (1992)
630635.
[41] T. Tatsuma, T. Watanabe, Model analysis of enzyme monolayer- and bilayermodied electrodes: the steady-state response, Anal. Chem. 64 (1992) 625
630.
[42] R.M. Ianniello, A.M. Yacynych, Immobilized enzyme chemically modied
electrode as an amperometric sensor, Anal. Chem. 53 (1981) 20902095.
[43] T. Yao, A chemically-modied enzyme membrane electrode as an
amperometric glucose sensor, Anal. Chim. Acta 148 (1983) 2733.
[44] Y. Okawa, H. Tsuzuki, S. Yoshida, T. Watanabe, Glucose sensor carrying
monomolecular layer covalently bound to tin(IV) oxide electrode, Anal. Sci. 5
(1989) 507512.
[45] H.R. Mahler, G. Hbscher, H. Baum, Studies on uricase: I. Preparation,
purication, and properties of a cuproprotein, J. Biol. Chem. 216 (1955) 625
642.
[46] M. Nanjo, G.G. Guilbault, Enzyme electrode sensing oxygen for uric acid in
serum and urine, Anal. Chem. 46 (1974) 17691772.
[47] K. Kahn, P. Serfozo, P.A. Tipton, Identication of the true product of the urate
oxidase reaction, J. Am. Chem. Soc. 119 (1997) 54355442.